Project Save Lives: Rapid treatment protocol using peer recovery specialists in the emergency department

Drug overdose deaths, including those involving prescription opioids and heroin, have more than quadrupled in the United States since 1999 [1]. Synthetic opioids, particularly illicitly manufactured fentanyl (IMF), have driven the largest increase in overdose deaths from 2014 to 2017 [2,3]. In 2015, both the DEA and the CDC issued nationwide alerts identifying illicitly manufactured fentanyl as a threat to public health and safety. In 2017, Florida heroin deaths had increased by 72% from 2014 (549 to 944) and fentanyl deaths had increased over Abstract

224% (538 to 1,743) during this time period [4,5]. In May 2017, the governor of Florida offi cially declared the opioid epidemic a public health emergency, and nationally the president declared a public health emergency in October, 2017.
North Florida, and Duval County (Jacksonville) in particular, has not been spared from this epidemic. In Northeast Florida, deaths caused by heroin-involved overdose rose from 25 in 2014 to 93 in 2017, a 272% increase [4,5]. Fentanyl overdose deaths (not including analogs) increased over 800% (33 to 302) from 2014 to 2017 [4,5]. Duval County is the area/county in Florida with the fourth highest heroin and fentanyl related deaths but has the highest incidence based on population size [5]. Calls While not clinical professionals, these specialists have personal experience in recovering from addiction and substance use disorders, allowing them to connect with opioid overdose victims in ways that health care professionals often are not able. In addition, overdose victims are more inclined to engage with individuals who have had the same experiences [6].
Peer based support services have been gaining national acceptance as an effective way to navigate the recovery process in substance use treatment programs [7] and increase treatment adherence [8]. The use of CRPS in the emergency department setting as a means to connect overdose patients to substance use treatment is less common and a fairly new concept in addressing the overdose epidemic; therefore, few studies have examined the effi cacy of these services [6,9,10].

Methods
This was an observational study as part of a program evaluation project. The evaluation approach includes a formative evaluation process which is intended to foster continuous improvement in the development and implementation of the project as well as a summative evaluation to assess the overall impact on the project in meeting its stated goals. Initially, two local hospitals within the same healthcare system volunteered to participate in the PSL program and were included in the evaluation.

Intervention
The intervention was developed by a community based, multi-disciplinary group that included substance use treatment providers, hospital ED staff, CRPS, and leadership from the local health department and the county's emergency medical services and was based on recommendations to address opioid overdoses from the Centers of Disease Control and Prevention (CDC) [11] and Substance Abuse and Mental Health Services Administration (SAMHSA) [12] as well as Rhode Island's Anchor Program eligibility requirements included that the patient be at least 18 years of age and have a positive response to naloxone (administered by EMS or ED staff) and/or positive fentanyl urine screening test [14]. After the ED staff stabilized the patient, the CRPS spoke with the patient regarding his/ her situation and the possible next steps, including treatment options for drug dependency including detox with MAT induction, residential or outpatient services. Once consent for the PSL program was obtained from the patient by the CRPS, the patient was then transported directly to a local substance use treatment facility from the ED. If a patient declined treatment, the CRPS offered naloxone and training on its use, a brochure on treatment options and asked if they can contact them in the future. If the patient agreed, the CRPS contacted the patient on a weekly basis for up to six weeks to see if he/she was now interested in pursuing treatment options. As the program matured, procedures were refi ned and additional treatment protocols were added. For example, it became apparent early in the program's implementation that patients experiencing withdrawal symptoms due to naloxone reversal were often too agitated to discuss their opioid dependency with the CRPS. The immediately transport the patient to detox or accompany him/ her to the outpatient treatment center the next day.

Study design
Overdose patients treated in the two participating EDs were categorized into two groups: those who were approached by a CRPS during their ED visit and those who did not have contact with a CRPS due to the staffi ng schedule (intake days versus non-intake days). The CRPS was only available in the

Evaluation outcome measures and data collection
The goal of Project Save Lives is to reduce opioid related overdoses and mortality.The primary outcome measures for the PSL initiative are the reduction of repeat overdoses and overdose related deaths within 6 months of initial overdose (index visit) treated in the ED at the participating hospitals.
Additional outcome measures were tracked on those patients who were approached by a CRPS during their ED visit include: 1) Linkage to substance use treatment services, and 2) Medication Assisted Treatment (MAT) among those who enter treatment services.
All ED overdose patients were given an identifi cation number and data were collected and entered in a secure database by the project coordinator who was authorized to access to the hospitals' electronic health record system. The data collected for all overdose cases for the index visit included: For Phase 2, data were collected on all overdose cases for 6 months after the index visit. The hospital's medical records were reviewed for subsequent overdoses and deaths. The PSL treatment data on the types of services that were accessed by these patients (detox, residential, outpatient and MAT) were also collected. Vital statistics from the state department of health were reviewed for any deaths that may be related to a drug overdose during the period of one year post index case for all patients (intake and non-intake days).

Analysis
The descriptive analysis examined the demographics of the patients during intake days and non-intake days to see if there were differences in these populations. In addition, the patients seen in the ED during intake days were further categorized to those who agreed to participate in the PSL program and those who declined. A comparison of overdose history (six months prior and 6 months after the index visit at the participating hospitals) of those who agreed to participate in the program, those who were approached by a CRPS but declined to participate and those who came to the ED during non-intake days was performed. In addition, an analysis of types of substance use treatment services received by the PSL participants was examined as well as if any deaths occurred during this time period. Analyses were performed using SAS 9.4 (SAS Institute, Cary NC).

Results
There were 45 individuals seen in the two participating EDs for an opioid related overdose on non-intake days (comparison group) and 105 individuals were seen on intake days (cases) from 11/16/2017 to 05/31/2018. Four of the 105 individuals died at the hospital and are excluded from the analysis. All individuals discharged from the hospital (146) were then followed for six (6) months to track their outcomes. The median age of those who arrived in the ED for an overdose on nonintake days was 36.3 while the median age of those who arrived on intake days was 36.4. White, non-Hispanic individuals who were uninsured or on Medicaid accounted for most of the overdoses on both intake and non-intake days ( Table 1). The only signifi cant difference between the two groups was gender where there were more males than females for those patients who arrived during the intake days.
Of the 101 patients who were discharged from the ED during the intake days, 32 (31.7%) patients received substance use treatment and 69 (68.3%) patients were offered but declined treatment. However, 66 of the 69 patients who declined treatment did agree to have a CRPS contact them post ED discharge. Of those who received treatment, eighteen (56.3%) participated in all three options -Detox, inpatient, and outpatient. Detox was the most used at 96.9% (31 people) and was highly recommended by the CRPS as the fi rst step for the MAT induction process but was not required if they had their fi rst dose of buprenorphine in the ED. Twenty-fi ve people (78.1%) went into residential treatment, and twentyfour people (75.0%) received MAT as part of their treatment process ( Table 2). None of the individuals who went to the ED during non-intake days went to the participating substance use treatment center during the 6 month follow-up period. Repeat overdose ED visits: Overall, overdoses treated in the ED declined with all patients during the 6 month period after the index visit as compared to the six months prior (13.0% had overdosed at least once before and only 5.5% treated again after the index visit) ( Table 3). A higher proportion of patients who went into substance use treatment had a repeat overdose as compared to those who did not get treatment. Of the 45 people who came to the ED for an overdose on a non-intake day, six (6) had been seen in the ED for a collective total of 12 previous overdose visits during the prior six months of their index visit (average 2.0 visits). In the 6 months after their ED index visit, the non-intake days group had two (2)

Discussion
There is a scarcity of evidence-based programs to effectively address opioid addiction on a large scale through a comprehensive program implemented in the ED. There have been numerous studies in the use of naloxone, both in the community and in the ED and its effectiveness in the reversal of respiratory depression [15][16][17][18] but that is only the fi rst step in addressing major harm from opioid addiction. Additionally, there have been a few studies that show brief intervention/ counseling of patients who self-report opioid addiction or misuse in the ED resulting in the reduction of opioid overdose risk behaviors [19][20][21]. A study by Yale School of Medicine showed that brief intervention/counseling of individuals screened for opioid abuse were more likely to engage in treatment and those who received both the brief intervention and initiation of buprenorphine in the ED had the highest rate of enrollment into treatment [22,23]. However, EDs do not typically intervene beyond acute medical stabilization or address the patient's addiction. Several surveys of Emergency Departments across the country show that the majority of EDs have limited policies in the management of opioid related encounters that go beyond medical stabilization [24,25]. A few communities have begun to tackle opioid overdoses using recovery peer support in the ED to refer patients into treatment programs including Opioid Overdose Recovery Program (OORP) in New Jersey and several hospitals in Rhode Island in conjunction with the Rhode Island Department of Health and a community recovery center [26,27]. These programs have shown promising results and only have begun tracking longterm outcomes from these interventions (28]. The Project Save Lives initiative is showing that aggressive linkage to substance use services in the ED using on-site recovery peer specialists is an effective program in addressing the opioid epidemic, especially in increasing access to and participation in substance use treatment programs. Of the 101 patients approached by the CRPS in the ED during the evaluation period (105 arrived during intake days minus 4 who died at the hospital), over 30% accessed some type of substance use

Limitations
This was an observational study and as such is subject to confounding, especially due to selection bias. We encountered several limitations in tracking patients outcomes six months post discharge from the ED index visit. PSL staff only had access to the ED records of the participating hospital system and therefore, it individuals were treated for an overdose in another hospital, we were not able to capture their data. We only had access to substance use treatment data from our participating provider; however, they are the largest treatment provider in the region. Additionally, if any of the individuals (both PSL participants and non-participants) had a fatal overdose outside our Medical Examiner's Offi ce 5 county service area, we were not able to retrieve this information.
Since the initial evaluation, PSL has expanded into several other hospital and free-standing EDs in the community with a goal to have all hospitals in the county participating by early 2020. A follow-up evaluation is being planned to see if the program is effective on a larger scale.

Conclusions
Even though EDs have been struggling with the increased volume of opioid related overdoses, they also provide a unique opportunity to address harm reduction and link patients to substance use treatment services. The PSL program resulted in increasing access to and participation in substance use treatment programs of overdose victims. Immediate engagement by a CRPS with direct linkage to substance use treatment services in the ED is promising to be an effective strategy in addressing the opioid epidemic, especially those who have experienced a nonfatal overdose. Additional research is needed to evaluate the long-term impact of ED engagement programs.

Ethics approval and consent to participate
This project was approved by the University of Florida Institutional Review Board.

Funding
This work was supported by an appropriation from the Jacksonville City Council, Jacksonville Florida.